Non-ribosomosal peptide (NRP) and polyketide (PK) metabolites synthesized by soil-dwelling bacteria are an important source of clinically useful compounds including many antibiotic, antitumor, and immunosuppressant compounds; however, the classic, culture-based approach to natural product isolation is unable to examine the metabolites of the 99% of bacteria that are recalcitrant to growth in laboratory conditions. The genomes of this uncultured majority represent one of the largest pools of unexplored biosynthetic diversity, a resource that can be evaluated for the production of useful metabolites by cloning environmentally-derived genomic DNA (eDNA) directly from soil and expressing the recovered gene clusters in heterologous hosts. While it would be desirable to screen eDNA libraries in a high-throughput manner, the large size of eDNA libraries (107 members) and the low abundance of metabolite-producing genes have made it difficult to do so in the past. This proposal aims to facilitate small molecule drug discovery efforts by developing a series of bacterial hosts that can be used to rapidly enrich eDNA libraries for clones containing biosynthetic clusters and then to screen those focused sublibraries for the production of antibacterial metabolites. Previous work has shown that genetic complementation of phosphopantetheinyl transferase (PPTase) deficient bacterial strains by eDNA libraries results in focused sublibraries that are richer in NRP/PK- encoding genes than any other published pool of DNA. Because the selective strategy is dependent upon complementation of NRP-encoded siderophore biosynthesis, a highly conserved iron-acquisition system, almost any bacterial strain can be transformed into an enrichment host. In Aim 1, the trainee will create a total of two Gram-negative and two Gram-positive enrichment hosts belonging to the Streptomyces, Burkholderia, and Pseudomonad families by using targeted gene deletions of the native PPTase and non-NRP-siderophore genes. In Aim2, the trainee will create eDNA libraries using shuttle vectors capable of replicating in either Gram-positive (pWeb436) or Gram-negative (pJWC1) bacterial hosts, aiming for the creation of two libraries per year, or four host-library pairs per year. Each library-host pair will be enriched for NRP/PK biosynthetic genes by selection on low-iron media. In Aim 3, the trainee will screen the libraries for antibiotic activity using overlay assays, and spot-assays with concentrated culture extracts. Clones exhibiting antibiotic activity will be isolated and used to characterize the antibacterial metabolites they encode.